By Annabel Harris*
On a windswept mountain retreat, a group of billionaire tech moguls gather, insulated from the world as it burns. This is the opening tableau of Mountainhead, a recent Hollywood film that feels less like science fiction and more like a warning shot. Their creation—a social media platform called Traam—has unleashed a torrent of AI-generated disinformation: deepfakes, fabricated news, and visuals so convincing that even the creators can no longer tell what’s real. As riots erupt and trust in society crumbles, the tech titans debate the authenticity of a grotesque video—a child juggling severed feet—while their AI filters fail to stem the chaos. The world outside descends into anarchy, and the line between truth and fiction dissolves.
It is a story that should be confined to the screen. But the uncomfortable truth is that Mountainhead is not far from our own reality. The same artificial intelligence that can conjure up fake wars and viral hoaxes is now being weaponized in ways that are even more sinister—and children are paying the price. According to the Internet Watch Foundation (IWF), AI is being used to create child sexual abuse imagery at a scale and speed never seen before. In just one month, over 20,000 AI-generated images were found on a single dark web forum, with more than 3,000 depicting criminal child sexual abuse. These aren’t crude cartoons or obvious fakes. The most sophisticated images are so realistic that even seasoned analysts struggle to distinguish them from real photographs.
Nowhere is the threat more acute than in Africa, where the digital revolution is outpacing the development of online safety measures. Africa is projected to add 337 million new internet users between 2024 and 2029—a 52% leap that will bring the total to 1.1 billion. One in three internet users on the continent is a child.
The continent’s young people are seizing the opportunities of the digital age, but the safety net beneath them is full of holes. The African Union’s Child Online Safety and Empowerment Policy, adopted in 2024, is a landmark step—the world’s first comprehensive regional framework for children’s rights in the digital environment. But the reality on the ground is that the rapid expansion of connectivity has left children exposed. Many lack the digital literacy to recognize online threats, and legal protections are patchy at best.
Police and prosecutors face a deluge of cybertips but lack the training, resources, and digital forensic tools to investigate them. Many officers have little experience with cybercrime, especially cases involving the dark web or cryptocurrencies. The result? Most cases go undetected and uninvestigated, and perpetrators operate with near impunity.
Foreign offenders have taken notice. In one chilling case, a retired French intelligence officer was arrested for orchestrating and paying for the sexual abuse of children in Kenya from afar, directing the crimes via encrypted apps. In another case, a German missionary posed as a church leader in Kenya, using his position to traffic and abuse boys as young as ten, filming the acts and distributing the clips overseas. The presiding magistrate condemned the pattern of “affluent Western and eastern visitors who hide their true selves in the church” to prey on Kenyan children.
But the story doesn’t have to end in despair. The same technology being weaponized against children can also be their shield. In a powerful example of AI for good, a graduate team from MTN, Africa’s largest telecoms company, recently won top honors at the UN Global Compact SDG Innovation Accelerator Programme for their AI-powered child online safety solution, YelloGuard. This system continuously scans digital environments for harmful content and predatory behavior, using contextual intelligence to detect not just obvious threats but also subtle forms of grooming and cyberbullying. When a risk is identified, it triggers real-time alerts to caregivers and moderators, enabling swift intervention.
What sets YelloGuard apart is its focus on Africa’s unique challenges—language diversity, local content, and region-specific threats. It’s not just a technical fix; it’s a proactive, adaptive approach that empowers children, parents, and educators to recognize and respond to online dangers.
The real breakthrough, though, lies in partnership. Africa’s telecom giants—MTN, Vodacom, Orange, Safaricom—reach hundreds of millions of users. The Africa Taskforce on Child Online Protection, launched with UNICEF, INTERPOL, and the Internet Watch Foundation, is showing what’s possible when public and private sectors join forces.
By embedding AI-powered safety features directly into the networks that children use every day, these partnerships can scale protection across the continent, reaching even the most remote communities.
As we mark World Children’s Day today, we face a choice. Will we allow AI to become the tool of chaos and exploitation? Or will we harness its power to build the shields our children so desperately need? Their digital future—and their safety—depend on it.
*Annabel Harris is a seasoned human rights and development leader who began her career at Amnesty International, managing the prestigious Human Rights Journalism Awards. As CEO of Reprieve, she steered the organization through rapid growth and high media visibility. She later led fundraising and communications for Jubilee Action, a children’s charity focused on global child welfare. Since 2011, Annabel has served as CEO of Network for Africa, championing mental health support for communities and children affected by conflict.
